Generate media right from your chat

Get your API key

Create a key in AI Media Studio under Workspace Settings → API. It starts with aims_.

Register the server (Claude Code)

One command registers the stdio MCP server for every project; start a new Claude Code session afterwards.

claude mcp add aims --scope user \
  -e AIMS_API_KEY=aims_... \
  -- npx -y @ai-media-studio/mcp

Or configure manually

For Claude Desktop, Cursor, Codex, or a project .mcp.json:

{
  "mcpServers": {
    "aims": {
      "command": "npx",
      "args": ["-y", "@ai-media-studio/mcp"],
      "env": { "AIMS_API_KEY": "aims_..." }
    }
  }
}

Generate

Just ask in natural language:

Use aims to generate a 16:9 image of a red panda
astronaut, then make a 6-second video panning across it.

For supplied image contents, call upload_image with raw image_base64 and an optional filename. For local files, run aims upload and use its returned URL in edit_image or generate_video. MCP uploads read no local files and require the media:write scope. Raster signatures are checked before upload. Failed or partial image batches return isError with all image results and completed URLs preserved.